Alzheimer's Disease (AD) is characterized by progressive cognitive and behavioral impairment. Over time, the patient is less able to care for him/herself, becoming more restricted to the home and the psychosocial environment of the caregiver (usually a spouse). As such, the patient's functioning may increasingly reflect the functioning of his or her spouse caregiver. Conversely, increased cognitive and behavioral decline in the patient may result in increased burden in the caregiver. This vicious cycle may lead to patient decline beyond that due to neuronal degeneration. At present, we know relatively little about how initial patient-caregiver dynamics affect long term outcomes. In our progress report we use a theoretical distress model to examine these issues. We demonstrate that caregiver risk (personality, health, distress) and resource (coping, social supports) variables at baseline predict subsequent caregiver burden, patient activities of daily living (ADLs), and nursing home placement. The proposed study will extend this work by interviewing our surviving couples twice (15-18 months apart) over the next 3 years. Although 71 couples are now available for follow-up, our power estimates have been evaluated downward from this figure because of expected attrition between now and the anticipated start date. Three hypotheses will be tested: 1) Caregivers who are most vulnerable to distress at baseline will report the most burden at follow-up; 2) AD patients living with caregivers who are the most vulnerable to distress and have the fewest resources at baseline, will have the lowest ADL levels at follow-up; 3) Caregivers who are more vulnerable to distress at baseline will, over time, admit their AD spouses into nursing homes at a higher rate than caregivers who are less vulnerable to distress. In each analysis, we will control for patient ADL levels and caregiver burden at baseline. Results of this research should have implications for public policy and intervention with caregivers of AD patients.